Artificial Neural Network-Based IoT Device for Predicting the Electromagnetic Compatibility of Medical Equipment: A Case Study of Hospitals in Rwanda

Electromagnetic radiation (EMR) and electromagnetic compatibility (EMC) issues in medical equipment can lead to inaccurate medical results, and harmful radiation emissions, posing risks to both patients and healthcare providers. Currently, there is a lack of effective prediction mechanisms to detect potential EMR and EMC issues in medical devices, contributing to equipment malfunctions and compromised safety. In this study, we propose an Internet of Things (IoT) device capable of predicting EMR and EMC in medical equipment using a Machine Learning (ML) model based on an Artificial Neural Network (ANN) model. The input parameters for prediction include temperature, aging, initial electromagnetic radiation (IEMR), frequency, and voltage of the device. The ANN model was trained on a dataset collected from four hospitals in Rwanda and its performance was compared with other machine learning algorithms, including Multilinear Regression, Decision Tree, Random Forest, and Support Vector Machine (SVM). The results demonstrated that the ANN model outperformed the other algorithms in prediction accuracy, achieving $99.5 \%$, thus demonstrating its suitability for real-time EMR and EMC prediction in medical devices. The proposed device exhibited reliable performance in a clinical environment, enhancing both the safety and functionality of medical equipment. It can also be employed by medical device manufacturers for testing and predicting the EMR and EMC reliability of their products. These findings highlight the potential of machine learning in improving the safety and accuracy of medical devices.